How does network analysis fit into the development of public administration as an academic discipline? This article tries to bridge theoretical developments in public administration between the middle of the 1970s and the first half of the 1990s. The benchmarks being used are (1) the now classical account by Vincent Ostrom of The Intellectual Crisis in (American) Public Administration - published 25 years ago this year - and (2) Christopher Hood’s reconstruction of core values in (British) new public management. Rather than representing analytical developments as an endless succession of different or even mutually exclusive approaches, this contribution tries to reconcile different foci for analysing public administration. Administrative theory provides us with a rather stable meta-theoretical framework for studying the meaning of quality in government, governance and public administration at various levels of analysis. By relating the emergence of network analysis in PA to parallel developments such as the resurgence of (neo-)managerial and (neo-)institutional analysis, it also becomes clear that network analysis is useful as an analytical device, but that it needs to be linked to theoretical perspectives that provide us with operational assumptions about ‘networks’. Network analysis in itself only provides split ground for reinventing government and refounding public administration.